Numerous AccuVote-TS units
configured in daisychain formation in the DeKalb County (amongst others)
in the Georgia elections received insufficient AC power, despite being
connected to AC power. Units not daisychained apparently
did not experience these problems.
[Ian S.
Piper] I don't think you have enough information to support your
conclusion. Although details are rather scarce immediately
after an election, it would be useful to get more info on the situation to
determine what really happened. Info such as,
- Where were the voting centers
where this problem occurred.
- What are the serial numbers
of the units that experienced the problem.
- Was the power connection from the booth to the unit
secure in each case?
- Were those cables were tested to ensure that they
were connected inside the
booth?
BTW: What is meant by "insufficient
AC power"?
Keith, would you be able to respond to Ian's
questions? I don't have the information he has asked
for.
- What effect would power cable length,
configuration, shielding, or any other cabling characteristics have on the integrity of power
delivery in the context
of daisychained
AccuVote-TS units?
[Ian S. Piper] Our power supply is a
wide ranging type that can operate on an AC voltage as low as
80Vac. The cable lengths used in elections wouldn't be an issue,
unless they went to ridiculous extremes. Personally, I'd
limit extension cord lengths to a total of 100 feet per daisy chain
and I'd use a heavy wire gauge extension
cord.
Why use a heavy wire gauge extension cord?
Could you be more specific as to power dissipation
in terms of cable length and cable gauge? Are you
saying that shielding is not an
issue?
- How much power is lost from the power
transferring through the AccuVote-TS unit itself?
[Ian S. Piper] There is no power lost; it is
consumed. Power is determined by voltage and
current. The voltage is constant between machines. Each
unit would consume anywhere from 0.4 Amps (Idle with a fully charged
battery) up to 1.5 Amps (Fast charging a flat battery).
- What is the ceiling number of units that may be
daisychained from one power outlet?
[Ian S.
Piper] The limit on current is 10 Amps as would be dictated by the
circuit breaker on the first booth of the daisy chain. As we don't
know how many units in a daisy chain have charged or
discharged batteries, we have to assume the worst, which is
that each unit is drawing the most current possible (1.5 Amps while
fast charging a flat battery). With a 10 Amp circuit breaker on
the first booth, it can supply 1.5 Amps to up to 6 units in a
chain. If the units aren't fast charging you could do more, but in
a situation where the AC power is out in the whole building
and the batteries are used to a flat condition, when the AC power
comes back up, you'd have all the units fast charging and drawing up to
1.5 Amps each. Therefore, you'd want to stick to the 6 unit limit
on a daisy chain.
NOTE: When the units are turned off and charging a battery,
the worst case is a 1 Amp current draw, so therefore 10 units can be
daisy chained in a warehouse charging
setting.
- What power cabling may be used, other than our
own designated ones?
[Ian S.
Piper] The customer can use extension cords and power bars to
deliver power to the units as long as they are rated to carry at
least 10 Amps. Our power cables are rated for 10
Amps.
- Is it the case that in order
to not overload the power drawn from a single power jack, the jack
should be connected exclusively to a power circuit, ie. the circuit
should service no more than that power jack. Would there be any
simple tool (other than a voltmeter) available that could determine that
a power jack is connected exclusively to a power circuit?
[Ian S.
Piper] You don't need to have a dedicated circuit for the outlet
that the daisy chain is attached to. Even if there are more
outlets on that circuit, the answer is "don't use them."
There's no
simple tool to determine that an outlet is isolated from others.
An electrician can make that determination using some tools that put a
frequency on the circuit and sensing which devices (lighting or other
outlets) share that circuit. Other than that, the old "turn
the breaker off and see what is not working" approach is simple enough
to determine what is on the circuit. Power
capacities should be part of the survey at each vote
center.